Description of Pipeline Strength Degradation as a Pure Death Markov Process and Its Implementation in IMP

The paper describes a new comprehensive Markov model of deterioration of the pipeline residual strength as related to its ability to withstand the inner pressure due to defects growth, and its implementation when designing the pipeline integrity management program (IMP). This Markov model is designed for defects that will cause pipeline rupture (the model describing pipeline leak is given in [8]). The pipeline degradation is described by a non-homogeneous pure death Markov process of the continuous time, discrete states type. The initial data used to construct this model is the whole set of defects found during an ILI, and verification or DA. This set is then filtered to select defects which will actually contribute to the pipeline rupture type of failure. For each of such defects its residual strength is calculated, using one of the three internationally recognized pipeline design codes: B31Gmod, DNV, and Shell-92. The constructed model is checked by an independent method, which compares the Markov model with real life experimental data, combined with sound analytical corrosion growth models. The above model takes into account formation of new defects and elimination of dangerous defects via pipeline repair. This approach provides an assessment of the conditional probability of rupture of a pipeline segment, as it is derived using the data set from a specific pipeline length. On the basis of this model an algorithm is constructed for optimizing the time of the next inspection, using the two-level maintenance policy. Results of numerical analysis show that the expected number and volume of repairs largely depend on the value of the ultimate permissible pipeline failure probability. A brief description of the software which implements the described above methodology is given.